118 Dr L. Mandl on the Scales of Fishes. 



the two last classes. At the same time, this memoir is of more 

 value than all the preceding works on the subject. 



Ehrenberg (Ann. de Poggendorf, vol. xxviii. Leipsic 1833, 

 p. 470) has lately described the crystals of silvery matter for- 

 merly observed by Reaumur. He likewise gives an analysis, 

 made by Henry Rose, of a small quantity of this substance. 

 It is soluble in diluted nitric acid ; the solution is not affected 

 by ammonia, and only very slightly so by the addition of oxa- 

 lic acid. The results of these experiments shew,— that this 

 matter evaporates by heat, without residuum ; that it contains 

 no lime ; that it is soluble in the acids, as well as in alcohol 

 and the alkalis (accompanied with ebullition in the latter). 



We shall afterwards give an account of our own experiments 

 on the subject of these crystals, made under the microscope. 



M. Agassiz, in recent times, has strongly drawn the atten- 

 tion of philosophers to the form of scales, by assuming them 

 as the basis of his classification. After having said a few 

 words, in the work quoted below, on the structure of the skin, 

 he speaks first of the position of scales, and then of their form. 



M. Agassiz explains the different forms of imbrication, with- 

 out entering into details on the disposition of the skin, and 

 continues thus : " It results from this that the position of the 

 scales varies much ; however, we can usually distinguish pretty 

 regular series, which enable us to determine their position 

 with accuracy, particularly in the case of imbricated scales. 

 The series are disposed obliquely from the fore part backwards, 

 from the middle of the back to the middle of the belly ; these 

 series may be called dorso-ventraJ. It is necessary further to 

 distingiTish the superior and inferior demi-series, and I would 

 call those which extend from the lateral line to the back me- 

 dio-dorsal, and di.stinguish the anterior and posterior medio- 

 dorsal, according as we wish to indicate those which are di- 

 rected from before backwards, or those which are inclined 

 from behind forwards. The same thing should be done with 

 the series below the lateral line, which I call medio-ventral, 

 .such as extend from the lateral line backwards or downwards, 

 being the posterior medio-ventral^ and those directed forwards 

 anterior medio-ventral. 



" Scales," continues M. Agassiz, " are contained in the mu- 



